CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   time step (https://www.cfd-online.com/Forums/main/4343-time-step.html)

yangqing January 30, 2002 10:32

time step
 
in fluent help, it says that the time step can be specified as you will,because it's method is steady. but i read some books, i feel the time step is relation to grid. why? thanks

wang January 30, 2002 11:13

Re: time step
 
There is a CFL number that constrain the time step. Generally speaking, the finer the grid is, the smaller the time step should be. Otherwise the solution may diverge. Furthurmore, there would be other constraint on time step concerning flow features. If you use RANS euqations, time step must be much more greater than the turbulent fluctuation time and smaller than the characteristic time of large scale.

Anup January 30, 2002 22:29

Re: time step
 
I do not know about Fluent, but some codes calculate both the convective(cfl number) as well as diffusive (von-neumann number) time step and then use the minimum of the two. So the time step varies with each iteration.

Anup.

John January 31, 2002 04:22

Re: time step
 
If one tries to obtain a steady solution by time marching, then there is no time-accuracy problem and the time step can be taken as big as allowable by solution stability.

If one simulates an unsteady flow, then Courant number is a good reference for choosing time step.


All times are GMT -4. The time now is 21:27.